How was Google Search Built? | Google Search Programming Languages and Tech Stack

how was Google built I thought this was
a good question but the answers I found
online weren't very satisfying so today
we're gonna do a deep dive how was
Google built stack shared I Oh tells us
that Google's using these technologies
here and this list is so broad that it's
kind of just like saying yeah Google
uses technology so if you like and
subscribe I'm just gonna open up with
the answer of what Google's tech stack
really is don't worry I'll give you a
second done all right
so please make sure to watch the end of
the video just so you know why we use
these technologies at Google I mean I
say we but I don't work at Google I
don't represent them in any way and to
be clear we're not talking about the
million products that Google has we're
just talking about Google search and
from my research I found like Google
search is made of the following tech
stack the front end is built in C which
is the front-end engine that spits out
HTML Javascript and CSS the back end is
primarily built with C++ with some
Python sprinkled in for the non
time-sensitive operations and the back
end database is built on C++ with their
proprietary BigTable software after
doing the research to come up with these
answers I feeling a lot more satisfied
because this makes a lot more sense than
some random JavaScript framework running
Google so for the front-end we have C
which for the longest time I could not
find what ran the front end of Google
search but in this very obscure web dev
subreddit post
I mean comment I found something that
looks like it might be gold here user
wisdom says you may be interested to
learn that at Google there is not an
HTML file for the home page in fact it's
a program written in C which is used to
generate it and his source is that he
worked there and was on the team
responsible for manually changing the
home page
Google Doodles which is pretty cool that
I'm reading something written by that
guy I couldn't find any more info on
this wisdom guy I checked out his github
it might be as github but might not be
but it's empty I mean for all we know he
could be lying but it sounds legit and
it's the most recent and definitive
answer I could find online so from my
knowledge how this would work is we have
the user ping google.com from their
browser Google servers would take the
info from your browser via like a cookie
I'm assuming and then they would feed
that information into a super
performance C program that C program
would create the JavaScript HTML CSS
file on the fly and return you a
personalized copy of the home page to
your browser and into your eyes for the
back end we have C++ and that's for the
search so it really seems like a mixed
bag of beliefs but from Wikipedia we can
gather that back rub
Google's predecessor was built using
Java and Python so we'll have to split
up the back end into three parts for
easy digestion it's gonna be the
crawlers for indexing the core PageRank
algorithm for sorting the index and the
query processing to parse the search
term and to find the result in the index
so starting with the crawlers I found
that they were initially built in Python
but they transitioned over to C++ the
next Google engineer explains here the
crawling is a relatively small part of
the building in index process Python was
abandoned around the year 2000 and
because of the bottleneck concurrency
they switched over to C++ in the fact
that the internet really hasn't changed
that much over the last decade so we
don't really Bennet they don't really
benefit that much from using Python for
it's faster development cycle so next
we're moving on to the core PageRank
algorithm and according to this
computation scientist at Google it's
written in C++ and using they also use
Python as a kind of glue and I'm
guessing that you can Python for like
non time-sensitive operations which
makes sense because they want their
search to be as fast as possible and for
the query processing where they actually
take the search term
and try to match it to one of the
indexes that they have I would think
that they are using something like a
ternary search tree for this because
this data structure supports additions
on-the-fly and it allows you to attach
arbitrary values to your keys which
makes me think of things like metadata
or something
I really wish there's more information
on this but it seems like it's all
really proprietary information and
they've kept it under wraps really well
so I couldn't really find a definitive
answer and now this brings us to the
database which is BigTable here we have
a video of Jeff Dean the lead of Google
ai explaining the goals of Google's
database which sums up why they had to
go with their own solution so their
goals were that they wanted to have an
asynchronous process to be continuously
updating different pieces of data they
needed to support very high read and
write rates with efficient scans and
joins and have the ability to examine
data changes over time so to solve this
they use a massive database called
BigTable which they built using their
Google file system or GFS the system
isn't fully relational but it's
apparently suiting their needs better
and based on a quick wiki search it's
written in C++
so just to recap Google search is built
on a tech stack of C for the front-end
C++ for the back-end server code with a
little bit of Python code and then
primarily C++ for their BigTable data
storage system hopefully that video was
good enough for like and subscribe if
you like this kind of Tuck content let
me know in the comment section below and
especially if you have another company
or something or product that you want to
know the tech stack of and you want me
to make a video on that let me know in
the comments section below this is code
phony out peace